Current drawing, photo editing and industrial design software requires manipulation with curve editing tools, which forces artists to work in terms of mathematical objects such as splines. Current 3D modeling software typically requires artists to understand, at least to some degree, the underlying mathematics and follow relatively complex technical procedures to create and manipulate 3D objects (e.g., 3D surface models).
Even for trained designers, it is a difficult task to draw a satisfactory freeform curve by applying only one pen stroke. Baudel, 7, 1994, “A Mark-Based Interaction Paradigm for Free-Hand Drawing,” Proceedings of UIST 1994, ACM Press, 185-192) showed a method to partially modify an existing curve by overstretching on it, Bae, S.-H., Kim, W.-S., and Kwon, E.-S., 2003. Digital Styling for Designers: Sketch Emulation In Computer Environment,” Lecture Notes in Computer Science (Proceedings of international Conference on Computational Science and its Applications 2003), 2689, 890-700 extracted a few simple element curves that automotive designers use generally, and suggested a method to get those curves as the weighted average of repetitive strokes.
To create a 3D curve from a 2D curve that the user draws, some assumptions or interface tricks are needed to infer the curve's depth information. The system of Cohen, J. M., Markosian, L., Zeleznik, B. C., Hughes, J. F., and Barzel, R., 1999, “An Interface for Sketching 3D Curves,” Proceedings of I3D 1999, ACM Press, 17-22, creates 3D curves by having the user draw a curve first and then its shadow. In some systems the drawing order is flipped. (Grossman, T., Balakrishnan, R., Kurfenbach, G., Fitzmaurice, G., Khan, A., and Buxton, B., 2002, “Creating Principal 3D Curves with Digital Tape Drawing,” Proceedings of CHI 2002, ACM Press, 121-128; Tsang, S., Balakrishnan, R., Singh, K., and Ranjan, A., 2004, “A Suggestive interface for Image Guided 3D Sketching,” Proceedings of CHI 2004, ACM Press, 591-598) The user draws a curve on an orthographic plane, and then draws a second curve on the ruled surface that is extruded in the normal direction to the orthographic plane. Bae, S.-H., Kijima, R., and Kim, W. S., 2003, “Digital Styling for Designers: 3D Plane-Symmetric Freeform Curve Creation Using Sketch Interface,” Lecture Notes in Computer Science (Proceedings of International Conference on Computational Science and its Applications 2003), 2889, 701-710, described a system to create symmetric 3D curves, if a user skilled in perspective drawing sketches a symmetric curve pair, the system samples 2D point pairs considering the vanishing point, and converts them Into 3D point pairs to reform a 3D point curve. Because most calculations are performed in 2D image space and the conversion from 2D to 3D is point-wise, the accuracy is not high and the resulting 3D point curve can be jagged. Karpenko, (X, Hughes, J. F., and Raskar, R., 2004, “Epipolar Methods for Multi-View Sketching,” Proceedings of the 2004 EUROGRAPHICS Workshop on Sketch-Based Interfaces and Modeling, 167-173, attempted a multi-view curve sketch. It is conceptually obvious, but not easy to use because an intermediate planar curve created from a curve sketched on a first view is difficult for the user to correlate to the second view.
Geometric modeling is a highly interactive task that requires the user to perform frequent navigation and operations on geometric and graphical objects. Current commercial 3D modeling software provides complicated user interfaces having menus and tool pallets. The user needs to have a good understanding of underlying mathematics and has to follow technical operational procedure. Zeleznik, R. C., Herndon, K. P., and Hughes, J. F. 1998, “SKETCH: An Interface for Sketching 3D Scenes,” Proceedings SIGGRAPH 96, ACM Press, 163-170 and Igarashi, T., Mafsuoka, S., and Tanaka, H., 1999, “Teddy: A Sketching Interface for 3D Freeform Design,” Proceedings of ACM SIGGRAPH 99, ACM Press, 409-418, showed a new possibility of simple and intuitive user interfaces for geometric modeling by introducing pen gestures.
The most basic problem in gesture-based user interfaces is to distinguish command strokes from draw strokes. In the many publications, the command and draw modes are separated to avoid ambiguities of stroke interpretation. A common way is to use pen barrel buttons, keyboard buttons, and specialized mode switches. Some studies employed pen gestures such as the double tap (Moran, T. P., Chiu, P., and Van Melle, W., 1997. “Pen-Based interaction Techniques for Organizing Material on an Electronic Whiteboard,” Proceedings of UIST 1997, ACM Press, 45-54); flick (Zeleznik, R. and Miller, T. 2006, “Fluid Inking: Augmenting the Medium of Free-Form Inking with Gestures,” Proceedings of the 2006 Conference on Graphics Interface, ACM Press, 155-162); and press and hold (Buxton, W., 1990, “A Three-State Model of Graphical Input,” Proceedings of IFIP Tc13 International Conference on Human-Computer Interaction, North-Holland, 449-456) Saund, E, and Lank, E., 2003, “Stylus input and Editing Without Prior Selection of Mode,” Proceedings of UIST 2003, ACM Press, 213-216, suggested inference and user mediation. Ramos, G., Boulos, M., and Balakrishnan, R., 2004, “Pressure Widgets,” Proceedings of CHI 2004, ACM Press, 487-494, and Tsang, S., Balakrishnan, R., Singh, K., and Ranjan, A., 2004, “A Suggestive Interface for Image Guided 3D Sketching,” Proceedings of CHI 2004, ACM Press, 591-598, explored using pressure sensing to distinguish between gesture strokes and draw strokes.
Considering the complexity of geometric modeling, it is not possible to operate every function of a modeling system using only pen gestures. Too many gestures and complicated gesture grammar may confuse the user without providing hints about how to interact with the computer. Gesture-based interfaces are inherently not self-disclosed unlike conventional graphical user interfaces known as WIMP GUIs. On-screen menus such as the marking menus (Kurtenbach, G. and Buxton, W., 1993, “The Limits of Expert Performance Using Hierarchic Marking Menus,” Proceedings of CHI 1993, ACM Press, 482-487), FlowMenu (Guimbretiére, F, and Winograd, T., 2000, “FlowMenu: Combining Command, Text, and Data Entry,” Proceedings of UIST 2000, ACM Press, 213-218), tracking menus (Fitzmaurice, G., Khan, A., Pieké, R., Buxton, B., and Kurtenbach, G., 2003, “Tracking Menus,” Proceedings of UIST 2003, ACM Press, 71-79) can be a complement to gesture-based interfaces.
Rather than “chicken scratch” style sketches, professional automotive designers express shapes with simple controllable curves having a good flow by “drawing from the shoulder.” Those curves generally have no inflection point or at most one. (Bae, S.-H., Kim, W.-S., and Kwon, E.-S., 2003, “Digital Styling for Designers: Sketch Emulation in Computer Environment,” Lecture Notes in Computer Science (Proceedings of International Conference on Computational Science and its Applications 2003), 2669, 690-700) They draw these lines very light at first, and then darken them up when the lines form a satisfactory shape. (Taylor, T., and Hallett, L., 1996, “How to Draw Cars Like a Pro.,” Motorbooks international) Curves having complicated convexity changes are created by connecting simple curves with smooth transitions. While sketching, they rotate paper from time to time to stay within an articulation comfort range. Drawing and paper rotation are subconsciously Integrated. (Fitzmaurice, G, W., Balakrishnan, R., Kurtenbach, G., and Buxton, B., 1999, “An Exploration into Supporting Artwork Orientation in the User Interface,” Proceedings of CHI 1999, ACM Press, 187-174)
Automotive designers take considerable care in setting up an initial perspective frame and continuously try to keep their sketches in correct perspective and proportion. Vanishing points, horizon lines, ground plane, center plane, box representing the outside dimensions of the car, ellipses with axle lines are used to establish a perspective view. (Robertson, S., With The Hot Wheels™ Designers, 2004, “How to Draw Cars the Hot Wheels™ Way,” MBI). Flipping paper and drawing on the back of the paper is a common technique to detect distortion and inaccuracy with a “fresh eye.” Some designers work with an underlay of a computer-generated perspective grid with wheels, 3D rendered image of existing cars, or package drawings where key dimensions and engineering hard points are indicated. On the other hand, they sometimes exaggerate and distort perspective and proportion intentionally to Increase the appeal of particular aspects of a design (they called it “cheating”).
Raster-type painting software allows designers to make the best use of their drawing skills. Furthermore, it provides many advantages that are not found in using traditional media including undoing actions, easy storing and retrieving results, working with unlimited layers and special digital effects. But, imperfect eye-hand coordination is a problem to some when using the standard graphics tablet devices that are not integrated with the display device physically. With no expense for undo, it increases the designer's tendency of expecting a “happy accident” of getting a satisfactory curve by just repeating a stroke and undo. Some software provides the canvas rotation mode that is invoked by pressing a button on the keyboard. However, most designers do not use the function and draw curves by twisting their head and body. As a result some designers feel they can not duplicate the quality of their paper strokes using a tablet.
In using painting programs, vector-type curves like NURBS curves are used as a supporting geometry that define snapping sweeps to control brushing or masks for brushing and gradation in a specified area. Creation of the snapping sweeps and masks require time-consuming, tedious manipulation of control points as in using vector illustration tools.